Oxford dons hope to cash in on next market crash

Click to follow
The Independent Online

A team of scientists from Oxford University claims to be able to predict stock market crashes with a model of trading behaviour that uses the laws of physics.

It is not the first time that academics have said they had fathomed the dynamics of financial markets and this latest attempt to predict the unpredictable may be met by scepticism. The contribution of academics to investing has been in doubt since the elaborate theories of two Nobel prizewinners led to the dramatic collapse of Long Term Capital Management, the hedge fund, in 1998.

Dr Neil Johnson and colleagues at Lincoln College hope to change that. Their model of the markets extrapolates from historic data on the Nasdaq using the principles of particle physics.

Most academic attempts to predict the market are a variation on theories about crowds, whether they contain people or inanimate particles. Details of the Oxford algorithm are a closely guarded secret, but it is known to assume that traders behave like several different kinds of robots. The members of each category respond to market events in broadly the same way, rather like particles in a physical system.

"We are trying to build the simplest model of market trades that includes just enough information to produce a realistic outcome," Dr Johnson said.

The model has identified the key warning sign of a stock market crash: everyone adopts the same investment strategy. In 1996-97, there was no identifiable pattern to Nasdaq trading. But in 1999, according to the researchers, patterns begin to emerge. Last year, it crashed.

"When there are pockets of predictability, there is a very high risk that co-ordination increases until it turns into panic," said Dr Johnson.

The model cannot yet predict further ahead than 20 days, and Dr Johnson declined the invitation to provide The Independent with a forecast. "I can't win. If I get it right I'll be inundated with inquiries," he said. "I'm not happy with media coverage."

The team will only share its wisdom after obtaining patent protection and Dr Johnson is already gearing up for commercialisation, having showcased the software to the likes of Goldman Sachs and Deutsche Bank. He believes that as many as one-fifth of the participants in a market could use the algorithm before that participation itself began to interfere.

Naturally, Dr Johnson says the work is not about the money, but his "pure scientific interest" in particle behaviour. But he confessed: "I'm a classic Essex boy: I'm a trader."

The City's investment strategists are sanguine about an attack from the ivory tower. Steve Russell, a strategist at HSBC, says the Oxford team is restating common sense. "Before crashes, people will always be doing the same thing," he said. "When Tony Dye [the fund manager who avoided technology shares] quit [Phillips & Drew] last year, the technology bubble burst within a week." It would, said Mr Russell, because Phillips & Drew was the last bastion holding out against tech shares.